STARDUST

Synthetic data for maritime autonomy

All-weather, all-sea-state perception data for autonomous vessels, USVs, and coastal surveillance. Generate the rare and dangerous scenarios you can never safely collect at sea, labeled to the pixel across EO, IR, SWIR, and radar.

BOOK A DEMO READ THE DOCS

Working model in one week. Zero sea trials.

The data that matters most is the data you cannot safely collect

Maritime AI fails for one reason above all: bad data. Real collection at sea is slow, costly, and weather-dependent, and it skews to calm, common, benign conditions. The cases that decide a mission, heavy sea states, night, fog, glint, crowded ports, and fast inbound small craft, are exactly the ones you cannot stage.

Stardust generates them on demand. Built on Unreal Engine and a learned rendering stack, it produces sensor-true, fully labeled maritime scenes across the full operational envelope, with EO, IR, SWIR, and physics-based X-band and S-band radar in registration. Closed-loop simulation then lets your autonomy stack act in the scene and be scored against ground truth, so perception and planning are validated before the first sea trial.

Why real data falls short

// collecting at sea today

  • ✕Collecting labeled data at sea is slow, costly, and weather-dependent
  • ✕The dangerous cases, heavy seas, night, fog, glint, and near-collisions, are exactly the ones you cannot stage
  • ✕EO alone misses targets that IR, SWIR, and radar would catch
  • ✕Open-loop data cannot validate autonomy: without closed-loop control, perception and planning failures stay hidden until deployment

Sensor-true data, perfectly labeled

Every sea state and conditionSea state, weather, time of day, sun angle, glint, and visibility across the full operational envelope.
Multi-sensor, registeredEO, IR, SWIR, and real-time X-band and S-band marine radar in perfect registration. Radar developed in partnership with Ansys.
Pixel-perfect labels and 3D metadataDetection, classification, and segmentation with ground truth on every frame, plus platform pose, object range, and class counts.
Closed-loop simulationStep through scenarios with a sim-in-the-loop API, replay real missions, and run CI batteries against your autonomy stack.
SENSOR COVERAGE
EOIR (MWIR/LWIR)SWIRX-BAND RADARS-BAND RADAR

EO, IR, SWIR, and physics-based X-band and S-band marine radar, in perfect registration and matched to your fusion stack.

Inside the maritime data

Neural-rendered sea Photorealistic harbor and open-water scenes
Sensor-true, with ground truth Live platform, environment, and objects-in-range metadata
Every modality, registered EO, segmentation, and depth from one scene
Closed-loop telemetry Multi-sensor mission replay in Rerun

What teams build with it

Collision avoidanceDetect and track vessels and obstacles in all conditions.
COLREGS scenario testingValidate behavior against ready-made scenarios that mimic the rules of the road.
Automatic target recognitionClassify vessels by type across EO and IR.
Vessel detection and classificationBoats, buoys, launches, and tugs, by range and class.
Maritime domain awarenessPersistent wide-area surveillance and tracking.
Port and harbor monitoringWatch busy, cluttered waterways around the clock.
USV and ASV patrolTrain and test autonomous surface vessel perception at scale.
Dark-vessel and small-craft detectionSurface low-signature and fast inbound contacts.
1 weekto a working perception model, zero fleet deployments
100×faster improvement with targeted synthetic data patches
2 NM / 250radar range and concurrent targets simulated, with ST Engineering
1,000+production-ready 3D maritime assets

Speaks your domain

The vocabulary, sensors, and benchmarks maritime teams actually use.

COLREGSARPAAISMDAUSVASVMUSVRHIBEOIRSWIRX-band radarS-band radarATRsea statedark vessellittoralsensor fusionsea clutterslant range

“We hit great real-world performance almost immediately. But even more impressive, Bifrost’s 3D metadata let us develop AI capabilities that just are not possible with real data.”

Perception lead, defense technology company valued over $30B

Trusted by the teams building autonomy on the water.

SaronicSeadronixHavoc

Questions teams ask

How do you train a USV to avoid collisions without sea trials?

Generate thousands of labeled collision and COLREGS scenarios across sea states, weather, and traffic, then close the loop so the autonomy stack acts and is scored against ground truth.

What sensors do you need for maritime ATR?

EO/RGB, thermal IR, and marine radar in X-band and S-band. Stardust generates all of them with pixel-perfect labels and 3D metadata.

Can you simulate marine radar?

Yes. Real-time, physics-based X-band and S-band radar with configurable carrier frequency, beamwidth, polarization, and antenna pattern, developed in partnership with Ansys and used for radar system validation.

How do you cover the long tail at sea?

Find the gaps and biases in your real data, then generate targeted synthetic data patches to fill them, around 100x faster than collecting more at sea.

What is closed-loop maritime simulation?

A step-through simulation where your autonomy stack observes synthetic sensor frames and issues controls each timestep, so perception and planning can be regression-tested in CI.

Unlock unlimited maritime datasets

Tell us what you are building and the scenarios you need. We will get you access.

BOOK A DEMO EXPLORE STARDUST